Update August 2025

This repo is an archive of the beyondplanck papers repository, copied from the now defunct gitlab. All the information below is out of date, but is left for historical reasons.

BeyondPlanck Papers

The full list of BP Papers (in PDF format) appear at:

https://papers.beyondplanck.science

Since these are unreleased papers, they are not publicly available and are behind a:

bps3cr3t. (note the period at the end)

NOT all PDFs are properly compiled yet (read on)

There is a Gitlab process that compiles all PDFs, once a commit to the repo takes place. This is done "automagically", nothing is required from your part.

General Instructions

How to make sure your paper is properly compiled:

• There is an execution pipeline that displays the current compilation tasks at: https://gitlab.com/BeyondPlanck/papers/pipelines (also accessible from CI/CD -> Pipelines menu)

• It displays a Status and the User that triggered it (more likely you)

• The status is clickable and it shows you all the executing Jobs.

• The Jobs are split in multiple 'paper_XX' jobs which are the latex compilation task for each paper (numbered as in HKE's overview paper)

• Each Job is appropriately colored (Green: ok, orange/red: faulty) and they are clickable.

• Clicking an a Job (more likely of your paper) shows you the latex compilation log, and potentially a useful error on why it failed.

• If your job fails to compile correctly, it will not be available in the https://papers.beyondplanck.science page (obviously)

• Fix the error and recommit (read Notes for help)

• The whole pipeline takes around 4 minutes to complete (as of now, with NOT all papers compiling)

Notes

• The papers are compiled with 'latexmk'. Actual command line is:

latexmk -cd -pdf -output-directory=../builds aDir/paper.tex -jobName=BP_XX

executed from the root of the 'papers' repo. Parameters:

• -cd: Changes to the given dir while compiling
• -pdf: Generates PDF
• -output-directory: Outputs ALL papers in a single dir (builds/)
• -jobName: Names each compiled paper as BP_XX

• Paper authors are NOT using a standardized way of compiling their papers. I made the assumption that 'latexmk' ought to compile them all (and it does in the 5-6 I have tried so far).

• You should be able to run the same command locally on your machine to see if it compiles and if it causes the same error as on Gitlab.

• If you would rather use an alternate way of compiling (although we SHOULD be as common as possible) let me know the command line to compile your paper differently and I'll change the compilation process

Proposal for Changes/Improvements

• Paper directories are hit and miss and very long (making script references difficult and very long). Can we all please, use:

  paper_XX

• Can we already use the final name of the PDF file? I do not know if there is a specific pattern that all Planck (or BeyondPlanck now) papers are using. Maybe "BP_Paper_title_in_camel_case"? As of now the names are all over the place, and I produce them all, temporarily for now, as BP_XX.pdf

• If you choose to help out and tidy things up (please DO), in order to do so without breaking the compilation pipeline, you will also have to change folder name and paper name in a couple of files too:

  1. .gitlab-ci.yml:

     • Change the relevant 'paper_xx' stanza with the new folder and paper name. (the 'aDir/paper.tex' and the 'jobName' from the command above)

     • If your paper_XX stanza is not there (I run out of steam creating them all) you can copy and paste an existing one (editing accordingly) and create your own.

     • If you DO add your own, add a dependency to the 'website' section too (you'll see it)

  2. website/index.html:

     • You can edit your Paper Title as it displays in the homepage and the name of the compiled PDF there (it's plain HTML, it should be self explanatory) For href=pdf_name, no folder required, where pdf_name should be the same as the jobName from the latexmk command above

That should be all...

It might be too much info but all of this is actually boilerplate code and only need to be set up once. Then it ought to work on it's own.

Do not hesitate to get in touch with me (@stratosgear gerakakis@planetek.gr) for any further clarifications.

Footnote:

The total time required for the papers to be compiled and uploaded to the website depends on the number of available Gitlab runners that do the processing. Gitlab offers a limited amount of computing minutes per project per month, that eventual will run out before each month's end. I have thus also added a runner that runs on a private personal server of mine, that should always by available to take the remaining load.

If anyone has access to an always on server, capable of running Docker, and willing to donate some spare cycles do let me know for instructions on how to join the Gitlab pool (hint: read gitlab-runner-readme.md)

A Gitlab runner is also able to execute locally on a laptop, even for a few hours per day, if your remember to start it. If even some of us were running one, that would be more than enough to cover our needs.